Czech J. Food Sci., 35, 2017 (1): 79–88 Food Technology and Economy, Engineering and Physical Properties doi: 10.17221/123/2016-CJFS

Chia and as Improvers of - Dough and Cookies

Ivan Švec*, Marie Hrušková and Barbora Babiaková

Department of and , Faculty of Food and Biochemical Technology, University of Chemistry and Technology Prague, Prague, Czech Republic *Corresponding author: [email protected]

Abstract

Švec I., Hrušková M., Babiaková B. (2017): Chia and teff as improvers of wheat-barley dough and cookies. Czech J. Food Sci., 35: 79–88.

Basic wheat-barley flour premixes were blended at ratios 70 : 30 and 50 : 50 (w/w), and white or dark types of whole- meal chia and teff replaced 5 or 10% of these bases. All non-traditional materials increased farinograph water absorption up to 10% in total, and with likely dilution of the dough gluten network, their addition increased the de- gree of dough softening up to twice. Physicomechanical properties of non-fermented dough were changing stepwise, mainly in terms of rising dough elasticity. Higher elasticity of wheat-barley dough was reflected in better both specific volume and spread ratio of cut-out cookies. Further recipe modification by wholemeal chia and teff flours enhanced these positive changes; cookie volumes based on the wheat-barley premix 50 : 50 were about 22% higher than their counterparts from the premix 70 : 30. For both attributes of cookie quality, somewhat higher values were observed for dark variants of wholemeal chia and teff. Moreover, the spread ratio of cookies containing teff types reached higher values compared to chia ones. Characteristic barley flour flavour could be less acceptable for common consumers; both wholemeal chia and teff flours were able to mask that by-taste.

Keywords: composite flour; non-traditional plant material; baking test; cookie; PCA

In human life, sweet taste belongs among others ents/10180/20562003/2701391501.xlsx/80b4a62a- to its pleasures. From a scientific point of view, 22d8-493b-9dc3-2c2648207246?version=1.1). For a sweet taste is classified among the four (or five) basic comparison, the European per capita consumption ones. Since the beginnings of civilisation, honey and of fine baked goods in the period 2005–2013 ranged matured fruits have been primary sources of carbo- from 7.52 kg to 8 kg (Statistica 2016). hydrates characterised by sweet taste, indicating ‘I’m Nowadays, at the time of increased interest of not poisonous’ and rendering energy to survive. From consumers in food composition as well as its health natural sources, industrial development contributed aspect, cookies represent a type of baked prod- cane and beet sugar and artificial sweeteners later. uct which could be easily adapted to such a trend. In connection with cereal raw materials, the confec- Cookies are a concentrated source of energy, mainly tionery art covering gingerbreads, cakes, biscuits, and owing to high sugar and contents. Dosage of for- cookies was progressively established. According to gotten of , sunflower, , chia, the Czech Statistical Office data, the average annual hemp, teff, and e.g. barley flour bring nutritionally consumption of preserved bakery products was 8.7 kg beneficial compounds such as proteins, unsaturated per capita in 2014 (ČSÚ 2015; www.czso.cz/docum fatty acids, or different forms of dietary fibre (Best

Supported by the Ministry of Agriculture of the Czech Republic, Project No. QI111 B053.

79 Food Technology and Economy, Engineering and Physical Properties Czech J. Food Sci., 35, 2017 (1): 79–88

doi: 10.17221/123/2016-CJFS

2009; Ohr 2009) and could lower energy intake at of dry and pre-hydrated chia flour on proper- the same time. On the other hand, the introduction ties; the latter chia form caused a lower diminishing of non-traditional raw materials and potential lower- of bread specific volume. Combined with fat, Coelho ing of sugar content have some technological limits. et al. (2015) recommended 7.8% of whole chia seeds Barley (Hordeum vulgare, the family ) be- and 10% of chia wholemeal flour for industrial use longs among basic cereals, traditionally used for because a decrease in bread specific volume was pearled barley, malt, and alcoholic drink production. still acceptable. On the other hand, the authors de- Barley β-glucans were acknowledged as nutritionally clared a benefit of such enhanced in a lower valuable dietary fibre (EFSA 2011); in mixtures with content of saturated fatty acids. For gluten-free diet, wheat flour, barley flour increases farinograph water Costantini et al. (2014) created bread based on absorption and prolongs dough development time as chia-Tartary blends (90 : 10 or 10 : 90); well as it causes a rise of water suspension viscosity final products were characterised by 20% of (Sullivan et al. 2010). Specific volume of bread and 74% of insoluble dietary fibre together with over containing 30% barley flour was about 25% lower 67% of α-linolenic acid. compared to the wheat control, and its flavour was Wheat biscuits containing 5–20% of chia flour had affected by typical barley taste (Hofmanová 2011). a higher nutritional value, mainly a higher amount Teff (Eragrostis tef, the family Poaceae) belongs to of protein, dietary fibre, antioxidants, and especially the main cereal produced in tropical Ethiopia. Very polyunsaturated fatty acids. Reversely, higher fat small seeds, coloured from white to red and dark content leads to accelerated lipid oxidation, indicat- brown, milled to wholemeal, are usually baked in a ing a shorter storage time (Mesías et al. 2016). Chia form of flat spontaneously fermented bread called wholemeal flour can also lower the specific volume ingera. A high portion of bran and germ present in of pound cakes (Luna Pizzaro et al. 2013); for this wholemeal brings nutritionally valuable compounds, type of cereal product, the negative effect could be such as high amounts of iron, calcium, and magne- limited by addition of hydrogenated vegetable fat. sium, able to cover the main portion of recommended Callejo et al. (2016) compared the effect of 15% dietary allowances (Hager et al. 2012). Based on a or 30% portion of red and white teff flour, added into large portion of prolamins, teff proteins are easily two wheat bread flour types of distinct baking quality. digestible. However, like in other cereals, a low con- Regardless of the tested recipe, such enhancement tent of lysine was reported (Adebowale et al. 2011). resulted in deterioration of principal parameters of Chia ( hispanica L., the family ) the bread structure, albeit dough rheological char- is an annual , typical of and acteristics were dependent on the teff variant and Central America. Nowadays Argentina, Columbia, level of its addition. The authors summarised that and Peru are the main producers. Seeds are rich in the teff variant must be chosen carefully to produce oil (30–33%) with important representation of un- good quality teff-based breads. Alaunyte et al. saturated linolenic and linoleic acids (Ciftci et al. (2012) verified their conclusion that a higher level of 2012) and naturally non-gluten proteins (16–26%, teff flour (up to 30%) worsened dough development Ayerza & Coates 2011). The composition also and increased crumb firmness and bitter flavour; the includes a high content of dietary fibre (37–41%) addition of selected enzymes may improve the loaf and phenolic compounds with antioxidant activity volume, crumb firmness and structure as well as (Reyes-Caudillo et al. 2008). overall acceptability of straight dough or sourdough As for final products, cereal researchers solved bread. Gluten-free breads, prepared by Campo et al. the application of chia and teff to recipes for bread, (2016), were based on flour supplemented with pastry (cookies, cakes), and pasta to a lesser extent. teff flour at ratios of 5, 10, or 20%, with or without The use of chia as a novel food is usually tested at three sourdough types. In terms of overall appearance, rates of 5–20%, although there exists a limitation by consumers preferred the variant containing 20% of teff the decision of the European Commission (EC 2013) flour, but in terms of flavour, bread combining 10% due to potential allergenicity and anti-effect to some teff and rice sourdough reached the best evaluation. hypertonic drugs. Verdú et al. (2015) replaced 5, 10, Among six variants of gluten-free cookies, the teff and 15% of wheat flour by chia flour, and stated the cookies did not overcome the product from coarse- slower bread staling due to higher water retention. grained rice flour, evaluated as the best in diameter Steffolani et al. (2015) compared the effect of 10% and spread factor, darker colour, and lower hardness 80 Czech J. Food Sci., 35, 2017 (1): 79–88 Food Technology and Economy, Engineering and Physical Properties doi: 10.17221/123/2016-CJFS

(Mancebo et al. 2015). Kenney et al. (2011) exam- according to the standard ČSN 560512-12:1995, and ined the effect of 25 and 50% replacement with teff the value reached 20.2 and 10% for both wholemeal chia flour on properties of gluten-free sugar and peanut and teff types, respectively. Similarly to the previous butter cookies. Teff decreased the spread ratio and paper (Hrušková & Švec 2015), composite flours hardness of cookies. Related to a control, the sen- included two groups based on cereal premixes C300 sory analysis did not demonstrate any difference in and C500 (WF/BF 70 : 30 and 50 : 50, respectively); in appearance, flavour, and taste of cookies containing both cases, wholemeal chia or teff flour substituted both proportions of teff. 5 or 10% of each premix in total (Table 1). The same Hager et al. (2013) prepared a set of gluten-free table summarises sample coding; prefix ‘C3xy’ and pasta; the sensory profile of variant was close ‘C5xy’ identifies composite flours based on premixes to the wheat standard, but teff reduced its quality. C300 or C500, respectively; for the tested wholemeal Determining in vitro digestibility, gluten-free teff spa- Ch1, Ch2, T1, and T2 flours, x is progressively equal ghetti had a lower glycaemic index than the control. to 1, 2, 3, and 4, respectively. Finally, y identifies the In the last several years, chia seeds have become very level of wholemeal flour addition: y = 1 means 5%, popular and their usage has spread from kitchens to while y = 2 means 10%. industrial production; not only bread but also crackers Flour and composite flour rheological properties. or cookies containing dark chia seeds can be bought Following the standards ČSN ISO 5530-1:1995, ČSN in groceries specialised in healthy nutrition. Also ISO 5530-2:1995, and ICC 126/1:1972, rheological oatmeal already belongs to industrial raw materials and pasting behaviour of non-fermented dough and for cookies manufacturing. Compared to chia and flour water suspension, respectively, was evalu- , teff and barley have a similar nutrition potential, ated using the Brabender apparatuses Farinograph, nevertheless they could still be categorised as non- Extensigraph, and Amylograph (Brabender GmbH, traditional materials. Such alternative materials are Germany). Following the ICC 173:2011 procedure, usually gluten-free, and evaluation of the rheological similar characteristics of composite flours were re- behaviour of composite dough based on wheat or corded on a Mixolab modern apparatus (Trippete & wheat-barley flour with addition of chia and teff may Renaud Chopin, France). Based on previous experi- facilitate their industrial usage. Farinograph, amylo- ence, among the farinograph features only the water graph, and Mixolab tests were used for this purpose; absorption and mixing tolerance index (degree of changes in viscoelastic properties of non-fermented dough softening) were used because these traits are composite dough were assessed by the extensigraph test. In cut-out cookies, the bakery potential of tested Table 1. Composition of tested flour blends tri-composite flour was quantified during a labora- tory baking test. The impact of the type and dosage of Flour, WF BF Non-traditional composite flour (g) (g)* flour (g)* non-traditional plant material was tested statistically, using analysis of variance and principal component WF 300 – – analysis (ANOVA and PCA, respectively). C300 210 90 – C3xy 200 85 15 C3xy 189 81 30 Material and methods C500 150 150 – C5xy 200 85 15 Flour types and composite flours. Both cereal flour C5xy 189 81 30 types, wheat flour (WF) and barley flour (BF), were from harvested in 2014. In WF and BF, protein WF – wheat flour; C300/C500 – wheat-barley composite flour contents were 11.98 and 9.40%, respectively. White and of 70 : 30 and 50 : 50 (w/w), respectively; BF – barley flour; dark chia seeds were bought in a specialised food shop, non-traditional wholemeals: chia white Ch1 (x = 1), chia and fine wholemeal flours Ch1 and Ch2 were prepared brown seed CH2 (x = 2); teff white seed T1 (x = 3), teff dark using a Concept KM 5001 blade grinder (seed dosage seeds T2 (x = 4); wholemeal addition levels: 5% (y = 1), 10% 50 g, operation time 1 min). A UK company supplied (y = 2); examples: C311 – composite flour based on C300, fine wholemeal teff flour T1 and T2, i.e. industrially containing 5% Ch1; C542 – composite flour based on C500, milled white and brown teff seeds. Protein content in containing 10% T2; *all amounts are percent of WF weight the tested alternative plant material was determined (i.e. enhancement levels 5 and 10%, respectively) 81 Food Technology and Economy, Engineering and Physical Properties Czech J. Food Sci., 35, 2017 (1): 79–88

doi: 10.17221/123/2016-CJFS sufficient to describe dough behaviour in relation to water absorption was caused by all three types of cookie quality. Going through laboratory practice, non-traditional materials; regardless of the dosage rheological measurement was carried out in one used and white/dark wholemeal types, the role of replication; repeatability of the tests was evaluated chia or teff was significantly more important than before, using an independent WF sample. Values of that of barley flour (Table 2). Barley, chia, and teff standard deviations are attached below data tables. flour contain a high portion of dietary fibre, which Cookies preparation. Baking trial was conducted is a known hydrocolloid. Rieder et al. (2012) con- in laboratory conditions, and the procedure was de- firmed the ability of BF to raise the water amount scribed in detail in a previous article (Hrušková & quantitatively for reaching demanded dough consis- Švec 2015). Briefly, composite flour was premixed tency. Water absorption increased by about 10–12% with sugar, NaHCO3, salt, and ascorbic acid using maximally, compared to values recorded for C300 a farinograph for 10 min (dough formula based on and C500 controls, owing to only 10% of chia or teff Kulp 1994). Then, sunflower oil was added, and added. A consequence of the higher water portion in distilled water was titrated until reaching the con- dough together with dilution of the supporting gluten sistency of 600 Brabender units (BU). Dough rested network was dough weakening after its overmixing. in a thermostat set to 32°C and relative humidity Barley flour did not contribute in this aspect at all, 95% for 30 min, and round cut-out cookies of 5 cm but in combination with chia, the degree of dough in diameter were formed manually. Baking without steaming was conducted in a laboratory oven pre- Table 2. Rheological characteristics of wheat flour and heated to 180°C, finished in 14 minutes. Quality tested composite flours evaluation of cookies was executed after 10 min Farinograph Amylograph cooling at ambient temperature. Specific volume Flour, composite WAF MTI T T AMA was evaluated on the basis of six samples, coupled in beg max flour (%) (BU) (°C) (°C) (BU) pairs; volume was determined by a rapeseed displace- WF 61.6a 55a 61.0d 91.0ef 680a ment method similarly to volume determination of b ab d ef b laboratory prepared bread (Alaunyte et al. 2012). C300 65.5 60 61.0 91.7 790 C311 70.6d 80bcd 60.2cd 94.7f 820c The spread ratio was calculated from cookie diameter C312 75.1fg 120ij 56.5b 90.3ef 900g and height, evaluated in triplicate with a calliper. C321 70.8d 80bcd 58.0bc 91.8ef 960i Quantification of the sensory quality of cookies C322 75.2fg 100fghi 56.5b 90.3ef 1000l was also described earlier (Hrušková & Švec 2015). C331 72.1e 85def 65.5ef 84.5bcd 990k Evaluated attributes are (surface) colour, aroma, taste, C332 75.9gh 100fghi 64.0e 82.0b 1000l consistency, stickiness during mastication, and overall C341 71.2de 115ij 64.0e 82.5b 860d acceptability. Each of the parameters could be scored C342 74.9fg 130j 67.0f 75.5a 950h by points 1–3, where the value 2 means an optimum; C500 67.1c 65abc 60.3cd 91.0ef 820c in fact, there are five categories for each sensory C511 72.1e 65abc 58.0bc 91.0ef 980j attribute because scoring 1.5 or 2.5 is also allowed. h ab a bc m C512 77.0 60 52.0 84.3 1020 Complex sensory score is then expressed as a sum of e a bc de i C521 72.2 50 58.0 89.5 960 all six parameters, with the value 12 meaning optimum. C522 77.0h 55a 56.5b 88.8cde 960i Data statistical treatment. Using the Statisti- C531 70.8d 90fgh 65.5ef 83.5b 900g ca 7.1 software (Statsoft, USA), differences in non- C532 74.9fg 90fgh 65.5ef 82.5b 1000l fermented dough behaviour and cookie properties C541 71.1de 105ghi 65.5ef 83.5b 870e were described by the HSD Tukey test (ANOVA, C542 74.6f 110hij 66.3ef 83.8bc 890f P = 95%). Comparison of the type of non-traditional Repeatability 0.2 4 0.5 0.9 4.2% plant material and addition level was analysed by principal component analysis (PCA). WF – wheat flour; C300/C500 – wheat-barley composite flour of 70 : 30 and 50 : 50 (w/w), respectively; WAF – water absorp- tion; MTI – mixing tolerance index (dough softening degree);

Results and discussion BU – Brabender unit; Tbeg, Tmax – temperature of gelatinisation beginning and maximum, respectively; AMA – amylograph a–l Flour and composite flour rheological behaviour. (viscosity) maximum; values in columns signed by different During farinograph testing, a verifiable increase in letters are statistically different (P = 95%) 82 Czech J. Food Sci., 35, 2017 (1): 79–88 Food Technology and Economy, Engineering and Physical Properties doi: 10.17221/123/2016-CJFS softening was determined to be approx. twice higher enhancement level (Figure 2). Extensigraph energy than that for both cereal premixes. In this regard, corresponds directly to the volume of baked product, behaviour of wheat-barley-chia systems differed which does not have a prior function in such pastry according to a barley flour portion. Data in Table 2 manufacturing compared to bread. allows presuming that not 30% but only 50% BF was Variance in amylograph features was comparable to a sufficient amount to hold water released from the the farinograph data; the impact of non-traditional protein structure during prolonged mixing, which material type and addition level factors intensified leads to significant dough strengthening during the from the beginning of gelatinisation to viscosity maxi- test. An opposite tendency of an increase in the degree mum. Tested materials supported the rise of viscos- of dough softening was confirmed by Alaunyte et ity maximum, and the level of chia and teff addition al. (2012) for wheat flour and WF-teff blend 90 : 10 could be considered as a dominant factor, although (61.7 and 108.3 BU, respectively). BF also contributed to this trend (from 680 BU to Figure 1 shows that all non-traditional raw materi- 490 and 820 BU for WF, C300, and C500 samples, als restricted non-fermented dough extensibility, and respectively; Table 2). On the other hand, hydrophilic- owing to this, an extensigraph ratio increased. For ity of wholemeal teff flour was lower during the RVA C300 and C500 samples, elasticity increased about test of wheat-teff blend 90 : 10; the peak viscosity was ca 40%, but extensibility diminished to 60 and 40%, statistically comparable to the WF control (595 vs. respectively. For WF-BF blend 60 : 40 (w/w), Rieder 588 mPa∙s; Alaunyte et al. 2012). Differences within et al. (2012) mentioned an extensibility decrease C300 and C500 groups were actually comparable; the to ca 65% of the wheat control. In wheat and teff range of measured values is equal to 200 BU. flour, the protein structure has a similar character; The Mixolab test allows evaluating both the pro- due to that, T1 and T2 wholemeal flours partially tein behaviour during mixing and polysaccharide alleviated the BF effect on dough extensibility. In all pasting properties (analogy of the farinograph and tested cases, extensigraph energy decreased about amylograph tests). Owing to the preparation of dough one-quarter and two-thirds, respectively, as 30% and at constant water absorption (59.4%), modification 50% of BF modified the dough composition; areas based on non-traditional material enhancement led of curves were generally smaller. The influence of to a significant increase in torque (parameter C1), both wholemeal flours could be depicted as minor, mainly for combination of WF-BF-teff (Table 3). causing diminishing in the extent depending on the Torque points C2, in which the heating input weakens

EEN (WBF30 group) EEN (WBF50 group) 120 ERA (WBF30 group) ERA (WBF50 group) 6

Sensorial score (points) ) % 8.0 10.0 12.0 14.0 ( 10 WBF50 group

80 4 +T2 ) WBF30 group 5 addition – ) ( 2

o and m i

c t 10 e ( a

p y R +T 1 g t y r 40 2 5

En e 10

+Ch 2 5 wholemeal

, 0 0 – 5 10 5 10 5 10 5 10 10 ou r fl 0 0 e +Ch 1 F 3 F 5 5 i t s B B

WF . +Ch1 +Ch2 +T1 +T2 o p W W . WBF50- Flour, composite flour, wholemeal addition (%) WBF30 Co m –60.0 –30.0 0.0 30.0 Sensory profile change (%) Figure 1. Effect of barley flour and chia or teff wholemeal addition on extensigraph test results; vertical lines identi- Figure 2. Effect of barley flour and chia or teff wholemeal fies 0.95 confidence interval; EEN, ERA – extensigraph addition on cookies sensory score; value 12 points is energy and elasticity-to-extensibility ratio; WF – wheat considered as optimum (i.e. reached by wheat cookies); flour; WBF30, WBF50 – wheat-barley flour premixes WBF30, WBF50 – wheat-barley flour premixes 70 : 30 and 70 : 30 and 50 : 50 (w/w), respectively; Ch1, Ch2, T1, T2 50 : 50 (w/w), respectively; Ch1, Ch2, T1, T2 – white and – white and dark chia/teff seeds wholemeal dark chia/teff seeds wholemeal 83 Food Technology and Economy, Engineering and Physical Properties Czech J. Food Sci., 35, 2017 (1): 79–88

doi: 10.17221/123/2016-CJFS protein functionality, confirmed a partial change in extent as in WF, C300, or C500 samples. In practice, protein structures. Within the C500 group, C2 median the baked product should be staling at a slower rate. was somewhat higher than within the C300 one (0.51 Evaluation of cut-out cookie quality. The recipe vs. 0.46 N·m, respectively); in these composite flours, water amount corresponded well with the farinograph a higher amount of barley β-glucans likely absorbed water absorption, i.e. BF at first and then wholemeal released water and thickened the dough. The pasting chia and teff flour increased the portion of added phase of the Mixolab test showed a course similar to water (about 14% compared to WF). Within the sub- the amylograph test, i.e. a dependence on the flour set of C500 samples, amounts of added water were blend composition, especially at torque points C4 comparable to C300 counterparts, although all these and C5 (stability of hot gel and retrogradation rate, values were lower than for C500 samples (59.7–64.1% respectively). The barley flour portion in composite vs. 68%; Table 3). In this regard, wholemeal teff flours flours did not play a decisive role; both wholemeal proved a higher capability than chia in both subsets. teff additions markedly shortened the stability of hot Similarly to the influence of hemp products on gel, and retrogradation did not occur to such a large cookie properties (Hrušková & Švec 2015), the specific volume of baked pastry was more significantly Table 3. Rheological characteristics of wheat flour and dependent on the tested recipe than the spread ratio; tested composite flours – Mixolab a partial dependence on extensigraph features could also be identified. Within the WF-BF-chia subset, Flour, C1* C2 C3 C4 C5 En- specific volume of cookies was supported by the composite ergy extensigraph ratio (r = 0.697, P = 95%), while the flour (N.m) PA volume of teff counterparts had a negative insig- WF 1.07 0.60 2.12a 1.92i 2.99o 138 nificant relation to extensigraph energy (r = –0.505, ab a gh k C300 1.17 0.51 2.24 1.69 2.56 126 P = 0.11333). Owing to quite high uncertainty of ab a efgh i C311 1.16 0.46 2.12 1.62 2.52 123 measurement (repeatability 0.3), ANOVA conjoined C312 1.17 0.46ab 2.12a 1.64fgh 2.58l 123 almost all samples of enhanced cookies into one C321 1.18 0.48ab 2.12a 1.67fgh 2.55j 125 group (Table 4). Within the C500 group, the aver- C322 1.16 0.48ab 2.14a 1.71h 2.61n 125 age of specific volumes was about ca 10% higher C331 1.27 0.48ab 2.18a 1.43ab 2.25f 118 than for C300 counterparts; in detail, there could be estimated an interaction of BF with wholemeal C332 1.23 0.44a 2.08a 1.40a 2.05a 111 chia and teff flours, stronger in the latter case (see ab a abcd c C341 1.17 0.46 2.20 1.47 2.15 115 the column ‘Group Mean’ in Table 4). According ab a abc b C342 1.18 0.45 2.19 1.45 2.14 114 to the spread ratio, teff samples based on the C500 C500 1.23 0.54ab 2.34a 1.72h 2.60m 131 premix (specimen foursome C531–C542) differed C511 1.16 0.53ab 2.23a 1.61defgh 2.58l 126 from the chia ones (foursome C511–C522); an average C512 1.13 0.51ab 2.18a 1.59cdefgh 2.52i 123 increase was 28 and 8%, respectively (as shown by the ab a defgh h ‘Group Mean’ column in Table 4). Finely ground chia C521 1.14 0.52 2.23 1.60 2.45 123 replacing 20% of wholemeal oat flour significantly C522 1.15 0.54ab 2.18a 1.65fgh 2.57k 125 increased both cookie diameter and height (about ab a bcdefg g C531 1.26 0.50 2.24 1.55 2.32 120 12.3 and 4.3%, respectively), but not the spread ratio ab a abcd d C532 1.23 0.47 2.14 1.47 2.18 114 (comparable to the wheat cookie control; Inglett et C541 1.21 0.51ab 2.30a 1.54abcdef 2.23e 120 al. 2014). Among 11 elaborations of gluten-free sugar C542 1.22 0.49ab 2.23a 1.49abcde 2.18d 117 snap cookies, Mancebo et al. (2015) highlighted Repeatability – 0.03 0.05 0.03 0.00 – buckwheat and teff ones due to a higher spread ratio, which corresponded with a higher protein content WF – wheat flour; C300/C500 – wheat-barley composite flour in their composition. of 70 : 30 and 50 : 50 (w/w), respectively; C1 – C5: points of Organoleptic profiles of cookies were changed the Mixolab curve; energy PA – energy equivalent to area slightly, the determined range was between 8.5 and under curve; *constant water addition (59.4%); a–ovalues in 13.5 points (optimum 12 points; Figure 2). The main columns signed by different letters are statistically differ- shifts occurred in colour (paler wheat-barley cook- ent (P = 95%); *all amounts are percent of WF weight (i.e. ies, dark dots on the surface when Ch2 was applied), enhancement levels 5 and 10%, respectively) taste, and consistency (data not shown); for sample 84 Czech J. Food Sci., 35, 2017 (1): 79–88 Food Technology and Economy, Engineering and Physical Properties doi: 10.17221/123/2016-CJFS

Table 4. Characteristics of baked cookies prepared from properties and cookie organoleptic profile, while composite flours PC3 to polysaccharide pasting properties and cookie profile, too. Comparing the explanation rate for SVB SPREAD Flour, cookie specific volume and spread (PC1–PC3 sum composite RWA group group mean mean 39 and 76%, respectively), a higher importance of flour mean mean the latter characteristic could be considered in WF 54.0a 160.1ab 160.1a 3.2a 3.2a b ab a ab a cookie quality description. Summarised, the PCA C300 56.3 155.8 155.8 4.1 4.1 method confirmed the effect of the three tested plant b a ab C311 57.3 152.5 4.3 raw materials and their dosages on the rheologi- c abcde ab C312 59.7 180.3 a 4.9 ab cal behaviour of non-fermented dough and cookie 167.6 4.4 C321 62.5fg 162.4abc 4.2ab characteristics. C322 60.1cd 175.5abcde 4.3ab Among dough and cookie quality features, five C331 63.4gh 169.3abcd 5.0ab main groups could be found in PC1 × PC2 area; C332 63.7h 159.7ab 4.1ab e.g. biopolymer absorption and pasting proper- 180.7ab 4.4ab C341 67.1i 207.9e 4.1ab ties (parameters MTI, TBE, RWA) or organoleptic C342 64.0h 186.2abcde 4.4ab attributes (Figure 3). Within the groups, known relationships were confirmed as well; depending C500 68.0i 160.0ab 160.0a 4.6ab 4.6ab on water absorption, dough consistency (C1) with C511 59.7c 174.5abcde 4.8ab de bcde ab diastatic power (C34) determines the specific of C512 61.0 190.0 5.0 nd ab b volume cookies (2 quadrant). In case of cookies, de cde 183.6 ab 4.7 C521 61.2 198.2 4.8 higher extensigraph energy (EEN) means lower cde abcde ab C522 60.7 171.8 4.3 specific volume; reversely, higher extensigraph C531 63.3gh 197.3cde 5.0ab ratio (ERA) allows better spread of such products. h cde b C532 64.1 198.0 b 5.4 c ef de 197.2 b 5.3 C541 61.8 202.9 5.1 –1.0 0.0 1.1 h bcde b 5 1.0 C542 64.0 190.7 5.7 C500 ERA Repeat- C3 0.2 6.5 0.3 ability C32 Spread C511

WF – wheat flour; C300/C500 – wheat-barley composite flour C2 C34 C541 SpecVol C312C512 C1 PA of 70 : 30 and 50 : 50 (w/w), respectively; RWA – recipe water C542C531 TMA Stickiness C521 C54 addition; SVB – specific volume of baked biscuits; SPREAD – WAF C522C300 a–i 0 C532 0.0 spread ratio (diameter-to-height); values in columns signed C12 AMA

PC2 (18%) PC2 TBEC331 by the different letter are statistically different (P = 95%) RWA MTIC341 C322-C311 C342 Consistency Taste EEN C312 (C300 + 10% Ch1), all these negative variants C321 WF C332 Overall were counted up. Acceptability of C300 and C500 acceptability Aroma controls was affected by typical barley flavour, for Colour which phenolic acids and proanthocyanidins are -5 -1.0 responsible (Holtekjølen et al. 2004). These chemi- -5 0 5 10 cal compounds cause bitter and astringent taste. PC1 (39%) Wholemeal chia and teff flours were able to supress Figure 3. Biplot of first two principal components (PC); that odour and aftertaste, but the higher recipe wa- non-traditional wholemeals: chia white seed Ch1 (x = 1), ter amount tempted to partial mouthful stickiness. chia brown seed Ch2 (x = 2); teff white seed T1 x( = 3), Statistical comparison of effects of non-tradi- teff dark seeds T2 x( = 4); wholemeal addition levels: 5% tional material type and addition level. The first (y = 1), 10% (y = 2); examples: C311 – composite flour three principal components (PC) explained 73% of based on C300, containing 5% Ch1; C542 – composite data variability; on average, PC1 accounted for 39%, flour based on C500, containing 10% T2; flour tri-com- PC2 18%, and PC3 16% (Table 5). The first PC is posites bases on C500 are written using italic, wheat- related mainly to dough rheological properties and barley-teff composites are differentiated by underlining; cookie characteristics, PC2 to protein mechanical for variable abbreviations (see Tables 2–4) 85 Food Technology and Economy, Engineering and Physical Properties Czech J. Food Sci., 35, 2017 (1): 79–88

doi: 10.17221/123/2016-CJFS

Table 5. Portion (%) of explained variability by the first three principal component (PC)

Test Feature PC1 PC2 PC3 Total water absorption 32* 0 55*** 87 Farinograph mixing tolerance index 49*** 6 1 57 temperature of gelatinisation beginning 43** 4 44** 91 Amylograph temperature of gelatinisation maximum 62*** 2 0 65 amylograph maximum 26* 0 62*** 88 extensigraph ratio 7 60*** 28* 95 Extensigraph extensigraph energy 54*** 14 16 84 C1 60*** 4 4 68 C2 53*** 11 4 67 C3 2 59*** 12 73 Mixolab C1–C2 76*** 0 0 76 (torque points and differences) C3–C2 34** 36** 6 76 C3–C4 80*** 8 1 89 C5–C4 85*** 2 3 90 energy PA 87*** 5 1 93 recipe water amount 77*** 6 8 92 Baking trial specific volume of cookie 30* 6 3 39 spread ratio 39** 32* 6 76 colour 0 62*** 1 62 aroma 1 51*** 1 52 taste 0 16 32* 49 Organoleptic profile consistency 1 14 39** 54 stickiness 33** 3 16 53 overall acceptability 13 31* 38** 83 Average 39 18 16 73

*, **, *** – correlation between dough or cookie quality feature and PC1, PC2, or PC3 – provable on P = 95%, 99%, and 99.9%, respectively; C1–C5 – points of the Mixolab curve; energy PA – energy equivalent to area under curve; underlined italic values signify a negative relationship Distribution of the tested samples is based on the in terms of rheological behaviour by farinograph, BF ratio (PC2) and type of non-traditional material in amylograph, extensigraph, and Mixolab. According composite flour (PC1) (Figure 3). Along the vertical to 2 × 8 recipe modifications, cookies were prepared PC2 axis, sweet dough machinability is improving on a laboratory scale. Barley flour increased water (decrease of extensigraph energy EEN and reversely absorption and so did both non-traditional materials an increase of ERA), resulting in a greater extent of (especially teff ones), in total about 10% absolutely. cookie spread. An increase of the cookie-making The higher water amount in dough together with potential of composite flours could be identified non-gluten protein introduction increased the degree along the PC1 axis; in this regard, both wholemeal of dough softening up to twice. Barley flour strength- teff flours showed a significantly higher contribution. ened dough resistance (elasticity) and reversely the extensibility; the extensigraph ratio increased approx. four times, allowing easier rolling out of dough and Conclusions shaping of cookies. Further change of dough machin- ability was observed mainly during testing composite Based on wheat-barley flour premixes 70 : 30 and flours containing wholemeal chia. The extensigraph 50 : 50, composite flours containing 5 or 10% of whole- energy (as an area below the curve) fell from 128.4 cm2 meal chia or teff white/brown flours were analysed to 38.7 cm2 owing to 50% supplement of barley flour 86 Czech J. Food Sci., 35, 2017 (1): 79–88 Food Technology and Economy, Engineering and Physical Properties doi: 10.17221/123/2016-CJFS for wheat flour, and to approx. 40.3 and 29.8 cm2 on Campo E., del Arco L., Urtasun L., Oria R., Ferrer-Mairal A. average by the addition of wholemeal chia and teff (2016): Impact of sourdough on sensory properties and flours, respectively. The hydrophilicity of chia and teff consumers’ preference of gluten-free breads enriched with was reflected in an increase of amylograph maxima up teff flour. Journal of Cereal Science, 67: 75–82. to the technical maximum of 1000 units (the values Coelho M.S., Salas-Mellado M.M. (2015): Effects of substi- of 680, 790, and 820 units for wheat flour and both tuting chia (Salvia hispanica L.) flour or seeds for wheat premixes, respectively), higher enhancement rate and flour on the quality of the bread. LWT-Food Science and higher amylograph viscosity. The Mixolab test revealed Technology, 60: 729–736. statistically verifiable differences during the heating Costantini L., Lukšič L., Molinari R., Kreft I., Bonafaccia G., and cooling phase of the test, indicating a dependence Manzi L., Merendino N. (2014): Development of gluten- of the rheological profile rather on composition and free bread using tartary buckwheat and chia flour rich in properties of polysaccharides than proteins. Regard- flavonoids and omega-3 fatty acids as ingredients. Food less of that, dough development of wheat-barley-teff Chemistry, 165: 232–240. composite flours was recorded earlier with sharper Ciftci O.N., Przybylski R., Rudzinska M. (2012): Lipid com- consistency maximum. Both wheat-barley cookie ponents of , perilla, and chia seeds. European Journal variants reached a similar evaluation as the control of Lipid Science and Technology, 114: 794–800. one. In correspondence to changed dough proper- EC (2013): Commission implementing decision of 22th Janu- ties, tri-composite flour cookie volumes and shapes ary 2013 authorising an extension of use of Chia (Salvia became greater as the enhancement level increased. In hispanica) seed as a novel food ingredient under Regula- cookie volume, a significant difference was registered tion (EC) No 258/97 of the European Parliament and of between wheat cookies, both wheat-barley controls the Council. OJEU. L 21/34–35. and wheat cookies containing 50% of barley flour EFSA (2011): Beta-glucans from oats and barley related together with teff flour (160.1, 155.8, 160 ml/100 g vs. health claims. EFSA Journal, 9: 2207. median 197.6 ml/100 g, respectively). Spread ratio of Hager A.S., Wolter A., Jacob F., Zannini E., Arendt E.K. (2012): enhanced samples also rose, from 4.6 for cookies from Nutritional properties and ultra-structure of commercial wheat-barley premix 50 : 50 up to 5.7 for the respec- gluten free flours from different botanical sources compared tive counterpart containing 10% of brown wholemeal to wheat flours. Journal of Cereal Science, 56: 239–247. teff. Each of the tested non-traditional plant materials Holtekjølen A.K., Kinitz C., Knutsen S.H. (2006): Flavanol and masked the barley characteristic flavour. bound phenolic acid contents in different barley varieties. Journal of Agricultural and Food Chemistry, 54: 2253–2260. References Hofmanová T. (2011): Characteristics of composite contain- ing domestic cereals. [Diploma thesis.] Prague, University Adebowale A.R.A., Emmambux M.N., Beukes M., Taylor of Chemistry and Technology Prague. J.R.N. (2011): Fractionation and characterization of teff Hrušková M., Švec I. (2015): Cookie making potential of proteins. Journal of Cereal Science, 54: 380–386. composite flour containing wheat, barley and hemp. Alaunyte I., Stojceska V., Plunkett A., Ainsworth P., Der- Czech Journal of Food Science, 33: 545–555. byshire E. (2012): Improving the quality of nutrient-rich Inglett G.E., Chen D., Liu S. (2014): Physical properties of teff (Eragrostis tef) breads by combination of enzymes in sugar cookies containing chia-oat composites. Journal straight dough and sourdough breadmaking. Journal of of the Science of Food and Agriculture, 94: 3226–3233. Cereal Science: 53, 192–197. Kenney E.S., Butler C., Moore C., Bhaduri S., Ghatak R., Ayerza R., Coates W. (2011): Protein content, oil content and Navde K.P. (2011): The effect of substituting teff flour fatty acid profiles as potential criteria to determine the in gluten-free sugar cookies and peanut butter cookies. origin of commercially grown chia (Salvia hispanica L.). Journal of the American Dietetic Association, 111: A63. Industrial Crops and Products, 34: 1366–1371. Kulp K. (ed.) (1994): Cookie Chemistry and Technology. Best D. (2009): Whole seed-better than whole grain? Cereal 1st Ed. Manhattan, American Institute of Baking: 280–283. Food World, 54: 226–228. Mancebo C.M., Picón J., Gómez M. (2015): Effect of flour Callejo M.J., Benavente E., Ezpeleta J.I., Laguna M.J., Carrillo properties on the quality characteristics of gluten free J.M., Rodríguez-Quijano M. (2016): Influence of teff vari- sugar-snap cookies. LWT-Food Science and Technology, ety and wheat flour strength on breadmaking properties 64: 264–269. of healthier teff-based breads. Journal of Cereal Science, Mesías M., Holgado F., Márquez-Ruiz G., Morales F.J. 68: 38–45. (2016): Risk/benefit considerations of a new formula- 87 Food Technology and Economy, Engineering and Physical Properties Czech J. Food Sci., 35, 2017 (1): 79–88

doi: 10.17221/123/2016-CJFS

tion of wheat-based biscuit supplemented with different Steffolani E., Martinez M.M., León A.E., Gómez M. (2015): amounts of chia flour. LWT-Food Science and Technol- Effect of pre-hydration of chia (Salvia hispanica L.), seeds ogy, 73: 528–535. and flour on the quality of wheat flour breads. LWT-Food Ohr L.M. (2009): Good-for-you-, Food Technology, Science and Technology, 61: 401–406. 63: 57–61. Sullivan P., O’Flaherty J., Brunton N., Erendt E., Gallagher Reyes-Caudillo E., Tecante A., Valdivia-Lopez M.A. (2008): E. (2010): Fundamental rheological and textural proper- Dietary fibre content and antioxidant activity of phenolic ties of doughs and breads produced from milled pearled compounds present in Mexican chia (Salvia hispanica L.) barley flour. European Food Research and Technology, seeds. Food Chemistry, 107: 656–663. 231: 414–453. Rieder A., Holtekjølen A.K., Sahlstrøm S., Moldestad A. Verdú S., Vásquez F., Ivorra E., Sánchez A.J., Barat J.M., (2012): Effect of barley and oat flour types and sour- Grau R. (2015): Physicochemical effects of chia (Sal- doughs on dough rheology and bread quality of composite via hispanica) seed flour on each wheat bread-making wheat bread. Journal of Cereal Science, 55: 44–52. process phase and product storage. Journal of Cereal Statistica (2016): Per capita consumption of fine baked Science, 65: 67–73. goods in Europe from 2005 to 2013 (in kilograms). Avail- Received: 2016–04–04 able at http://www.statista.com/statistics/265881/per- Accepted after corrections: 2016–12–23 capita-consumption-of-fine-baked-goods-in-europe/ (accessed March 22, 2016).

88